- Social media platforms must label AI-generated content clearly and prominently, the government has said
- Platforms cannot allow removal or suppression of AI labels once applied
- Automated tools must prevent illegal or misleading AI content on social media, the government said
The government has made it mandatory for social media platforms to place clear and prominent labels on AI-generated content. All such synthetic content must include identifiers, it said. A new government order said if a social media platform becomes aware of illegal or misleading AI content, it will be mandatory to remove or block access to it within three hours. Previously, this time limit was 36 hours.
Once an AI label or associated metadata is applied, social media platforms cannot allow it to be removed or suppressed, the order said.
The government has also ordered social media platforms to deploy automated tools to prevent illegal, sexually exploitative, misleading, or fraudulent AI content. The platforms must also warn users about the penalties for violating these rules at least once every three months.
The order is based on key amendments made in the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021. The amendments take effect on February 20.
The updated law increases the responsibility of social media companies and digital platforms regarding deepfakes and AI-generated content.
According to the government notification, "synthetically generated information" is now clearly defined. This includes audio, video, photos, or other visual content that has been artificially or algorithmically created and that represents a person or event as real.
The government has, however, clarified that basic editing, color correction, technical quality enhancement, translation, or document preparation will not fall under the category of synthetic content, provided it does not create a misleading or fake record.
Major social media platforms will also have to get a declaration from users confirming whether the content being shared is AI-generated. Companies will be required to verify this information through technical means. Non-compliance could lead to the platform losing its legal protection.
